AI in the CFO Office: How Finance Teams Get Real Value from NetSuite AI

AI can speed up finance work dramatically, but only when it is applied to the right processes. In Staria’s AI in the CFO Office webinar, Sales Executive Ari Karvola and NetSuite Solution Architect Kåre Kaasamoen explain how CFOs and finance leaders can tell where AI creates real value, where traditional ERP automation works better, and which AI capabilities are already available in NetSuite today.

With AI at the peak of inflated expectations on Gartner’s Hype Cycle, most finance teams have seen impressive demos. The harder question is what AI in the CFO office should actually be used for. This article summarises the webinar’s answer.

Key takeaways

• AI creates the most value in complex, interdependent and explorative finance processes, such as the month-end close, board reporting and cash forecasting.

• For simple, rule-based processes, scripts and standard ERP automation are cheaper, faster and more consistent than AI.

• NetSuite customers can use AI today: machine learning close tools, around 100 ready-made prompts and an MCP connection to AI assistants such as Claude and ChatGPT.

Watch the full webinar recording to see the live NetSuite demos.

Why AI won’t fix a broken finance process

In the CFO office, AI creates value only where your business processes do. Kåre, who has led ERP projects for almost 19 years, puts it simply: if processes are poorly defined or the data is weak, AI won’t fix that. Before bringing in AI, you need to know who owns the process, how it runs and what data it relies on.

That value also has to be measurable. Ari offered a rule of thumb: every euro you put into AI should pay itself back every year. Many AI projects fail because nobody defined which process the investment was meant to improve.

When is AI the right tool for finance? Five questions to ask

Kåre’s five-question framework sorts CFO office processes by the tool that creates the most value. The more complex and unpredictable the process, the more AI adds.

Not every process needs AI. Kåre described an accounts payable automation he built four years ago on simple accounting and approval rules: about 90 percent of vendor bills still go straight to posting. For predictable work like this, a script is cheaper, faster and gives the same result every time.

AI earns its place further down the list. Machine learning catches missing tasks and anomalies in the month-end close, while generative and agentic AI help controllers interpret numbers, prepare board material and explore questions where the answer isn’t known in advance.